public inbox for netdev@vger.kernel.org
 help / color / mirror / Atom feed
* [PATCH net-next] net: fclone allocation small optimization
@ 2026-01-16 16:44 Eric Dumazet
  2026-01-21  2:40 ` patchwork-bot+netdevbpf
  0 siblings, 1 reply; 2+ messages in thread
From: Eric Dumazet @ 2026-01-16 16:44 UTC (permalink / raw)
  To: David S . Miller, Jakub Kicinski, Paolo Abeni
  Cc: Simon Horman, netdev, eric.dumazet, Eric Dumazet

After skb allocation, initial skb->fclone value is 0 (SKB_FCLONE_UNAVAILABLE)

We can replace one RMW sequence with a single OR instruction.

	movzbl 0x7e(%r13),%eax // skb->fclone = SKB_FCLONE_ORIG;
	and    $0xf3,%al
	or     $0x4,%al
	mov    %al,0x7e(%r13)
->
	or     $0x4,0x7e(%r13) // skb->fclone |= SKB_FCLONE_ORIG;

Signed-off-by: Eric Dumazet <edumazet@google.com>
---
 net/core/skbuff.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)

diff --git a/net/core/skbuff.c b/net/core/skbuff.c
index f2367818c130bcd3095e9bd78fbc9f9a0ac83d4b..1c2b9ae7bce6457395a82d5bd577a9e125eceb95 100644
--- a/net/core/skbuff.c
+++ b/net/core/skbuff.c
@@ -723,7 +723,14 @@ struct sk_buff *__alloc_skb(unsigned int size, gfp_t gfp_mask,
 
 		fclones = container_of(skb, struct sk_buff_fclones, skb1);
 
-		skb->fclone = SKB_FCLONE_ORIG;
+		/* skb->fclone is a 2bits field.
+		 * Replace expensive RMW (skb->fclone = SKB_FCLONE_ORIG)
+		 * with a single OR.
+		 */
+		BUILD_BUG_ON(SKB_FCLONE_UNAVAILABLE != 0);
+		DEBUG_NET_WARN_ON_ONCE(skb->fclone != SKB_FCLONE_UNAVAILABLE);
+		skb->fclone |= SKB_FCLONE_ORIG;
+
 		refcount_set(&fclones->fclone_ref, 1);
 	}
 
-- 
2.52.0.457.g6b5491de43-goog


^ permalink raw reply related	[flat|nested] 2+ messages in thread

* Re: [PATCH net-next] net: fclone allocation small optimization
  2026-01-16 16:44 [PATCH net-next] net: fclone allocation small optimization Eric Dumazet
@ 2026-01-21  2:40 ` patchwork-bot+netdevbpf
  0 siblings, 0 replies; 2+ messages in thread
From: patchwork-bot+netdevbpf @ 2026-01-21  2:40 UTC (permalink / raw)
  To: Eric Dumazet; +Cc: davem, kuba, pabeni, horms, netdev, eric.dumazet

Hello:

This patch was applied to netdev/net-next.git (main)
by Jakub Kicinski <kuba@kernel.org>:

On Fri, 16 Jan 2026 16:44:02 +0000 you wrote:
> After skb allocation, initial skb->fclone value is 0 (SKB_FCLONE_UNAVAILABLE)
> 
> We can replace one RMW sequence with a single OR instruction.
> 
> 	movzbl 0x7e(%r13),%eax // skb->fclone = SKB_FCLONE_ORIG;
> 	and    $0xf3,%al
> 	or     $0x4,%al
> 	mov    %al,0x7e(%r13)
> ->
> 	or     $0x4,0x7e(%r13) // skb->fclone |= SKB_FCLONE_ORIG;
> 
> [...]

Here is the summary with links:
  - [net-next] net: fclone allocation small optimization
    https://git.kernel.org/netdev/net-next/c/79bfa5fb8587

You are awesome, thank you!
-- 
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html



^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2026-01-21  2:40 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-01-16 16:44 [PATCH net-next] net: fclone allocation small optimization Eric Dumazet
2026-01-21  2:40 ` patchwork-bot+netdevbpf

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox